Impacts of Biogenic Emissions Estimates from BEIS-3 on ...€¦ · Impacts of Biogenic Emissions Estimates from BEIS-3 MCNC The Impacts of Biogenic Emissions Estimates from BEIS-3

___________________Environmental Modeling Center

Impacts of Biogenic Emissions Estimates from BEIS-3

MCNC

The Impacts of Biogenic Emissions Estimates from BEIS-3

on Ozone Modeling in the Southeastern US

presented at the

11th Annual Emissions Inventory Conference, Atlanta, GA

April 15-18, 2002

Saravanan Arunachalam, Zachariah Adelman,

Rohit Mathur, & Jeffrey VukovichMCNC-Environmental Modeling Center

Research Triangle Park, NC [email protected]

2



MCNC

Outline

• Introduction• Background

• Objectives

• Modeling Systems

• Results

• Discussion

• Future Work

3



MCNC

Backgroundl Emissions of NOx and VOCs from anthropogenic (man-

made) and biogenic (natural) sources contribute to form ozone

l Grid-based photochemical air quality models used to study ozone formation – used for regulatory as well as scientific purposes

– use emissions processing systems to estimate emissions from various sources

l Large uncertainties remain in these emissions estimates– A factor of three for total VOCs nationwide is quoted in literature

l Understanding contribution of natural sources to ozone formation has significant implications towards choice and extent of controls needed on anthropogenic sources to demonstrate attainment of ozone standards

4



MCNC

Objectives

l To evaluate the impacts of using biogenic emissions estimates from a prototype version of BEIS-3 on ozone predictions

– Perform sensitivity simulations of MAQSIP and compare model predictions when using BEIS-2 versus BEIS-3 emissions estimates in a nested domain in the Southeastern United States

5



MCNC

Modeling Systems

l Fifth-Generation Penn State/NCAR Mesoscale Model Version 2 (MM5 V2.12)

l Sparse Matrix Operator Kernel Emissions (SMOKE) for generating emissions inputs

l Biogenic Emissions Inventory System – (BEIS-2 and BEIS-3)

l Multiscale Air Quality SImulation Platform (MAQSIP)

l Platform Independent I/O API format for all inputs and outputs

l June 19-30, 1996 episode in North Carolina

6



MCNC

Biogenic Emissions Inventory System

l BEIS-2– Five CB-4 Species (NO, ALD2, ISOP, OLE, PAR)

l BEIS-3– Prototype version with explicit speciation

– 1-km resolution landuse data from BELD3

– 230 different landuse types

– 34 BEIS-3 species mapped to 12 CB-4 Species (CO, NO, ALD2, ETH, ETOH, FORM, ISOP, MEOH, OLE, PAR, TOL, XYL)

– Treats CO, NO production from soils, and biomass burning

7



MCNC

The Multiscale Air Quality SImulation Platform

A comprehensive urban to intercontinental atmospheric chemistry/transport model

l Developed at MCNC in collaboration with EPA– Served as a prototype for EPA’s Models-3/CMAQ modeling system

l Has been used at various scales to study problems related to tropospheric O3, acidic substances, and aerosol formation

– Seasonal simulation for 1995 over Eastern US– Regulatory episodic applications in NC, VA– Other studies to analyze

• Relationship between pollutant transport and distribution over western Europe

• CO transport and distribution over the North Atlantic region andthe Asia Monsoon Region

• Transport and chemical evolution of Asian outflow over the Pacific Ocean

– Real-time Ozone forecasting system over the Eastern United States• Fine grids around Houston, Boston, Birmingham in summers of

1999-2001

8



MCNC

MAQSIP (..contd.)

l Modular– Physical/chemical processes cast as modules; each process module

operates on a common concentration field

l Flexible– Alternate process representations and algorithms– Provides a test-bed to compare different process

representations/schemes

l Expandable– New modules can be added to the system

l Multiscale– Multiple multi-level nested grids– One-way and two-way

l Generalized Coordinates facilitates– Use of different map projections – applications of various geographic

regions around the globe– Interfacing with meteorological models

9



MCNC

MAQSIP Processes Represented & Interrelationships

Horiz.

Adv

Vert

Adv

VertDiff

Horiz.

Diff

Eddy

Diff

Conc Init

CloudProcs

AerosolsGas-PhaseChem

BC Data

PhotJ-Table

Gas/AerosolIC File

Emission

Met Data

10



MCNC

Nested Modeling Domain[36/12/4-km Grids]

11



MCNC

AQ Monitors in the 12-km Grid

12



MCNC

Biogenic Emissions Estimates [Jun 24, 1996]

13



MCNC

Total (Anth+Biog) Emissions Estimates [Jun 24, 1996]

14



MCNC

Biogenic Emissions Estimates [36-km Grid: Jun 24, 1996]

15



MCNC

Biogenic Emissions Estimates [12-km Grid: Jun 24, 1996]

16



MCNC

Total (Anth+Biog) Emissions Estimates [36-km Grid: Jun 24, 1996]

17



MCNC

Total (Anth+Biog) Emissions Estimates [12-km Grid: Jun 24, 1996]

18



MCNC

Discussion

l Biogenic VOCs using BEIS-3 (compared to BEIS-2) increased by 79%, 62%, and 47% in the 36,12, and 4-km grids

l Total (Anth+Biog) VOCs using BEIS-3 (compared to BEIS-2) increased by 58%, 52%, 40% in the 36,12, and 4-km grids

l Corresponding increase in ISOP by 17%, 18% and 4%

l Large increases in MEOH, ETOH, and FORMl ALD2, PAR, and NO emissions decrease

– Total decrease in NO budget range from 1-3% in the 3 grids

19



MCNC

Modifications to MAQSIP

l Added two reactions to explicitly treat MEOH and ETOH oxidation within MAQSIP

– MeOH + OH àà HO2 + HCHO EtOH + OH àà HO2 + ALD2

– Reaction Rate Constants from DeMore et al. Chemical Kinetics and Photochemical Data for Use in Stratospheric Modeling, Jet Propulsion Laboratory, California Institute of Technology, 1997

l Dry deposition velocities for paraffins (PAR) were used as approximations for MEOH and ETOH

20



MCNC

Model Performance Evaluation [O3 Obs Threshold = 5ppb]

21



MCNC

Statistical Measures[Layer 1 O3 in the 36-km Grid: Jun 19-30, 1996]

0.714-0.5490.359.246.555.65BEIS-3

0.683-0.4830.336.846.553.25BEIS-2

0.249-0.0560.232.162.764.840BEIS-3

0.244-0.0100.22-0.762.762.040BEIS-2

0.1810.0200.14-2.174.872.760BEIS-3

0.1840.0610.13-5.274.869.660BEIS-2

GrErNBiasR2BiasAvg

Obs

Avg

Mod

Thres

(ppb)

AQM

22



MCNC

Statistical Measures[Layer 1 O3 in the 12-km Grid: Jun 19-30, 1996]

0.732-0.5770.359.547.957.35BEIS-3

0.700-0.5250.357.547.955.45BEIS-2

0.236-0.0560.242.063.865.840BEIS-3

0.224-0.0200.25-0.363.863.640BEIS-2

0.1760.0240.14-2.575.673.060BEIS-3

0.1710.0570.13-5.075.670.660BEIS-2

GrErNBiasR2BiasAvg

Obs

Avg

Mod

Thres

(ppb)

AQM

23



MCNC

Daily Max O3 in 12-km grid [Jun 21-30, 1996]

24



MCNC

8-h AQI Counts from BEIS-2 Simulation

25



MCNC

8-h AQI Counts from BEIS-3 Simulation

26



MCNC

27



MCNC

28



MCNC

29



MCNC

30



MCNC

Discussion

l Sensitivity study performed to compare BEIS-2 vs. BEIS-3 emissions estimates in MAQSIP

– Biogenic VOCs increase due to large amounts of MEOH from BEIS-3

– MAQSIP CB-4 chemistry updated to treat explicit alcohol oxidation

– Average 1-h O3 increases by 1-3 ppb in all grids– Leads to better model performance at a 60ppb threshold– Some correlation seen in predicted ozone increases due to

addition of radical sources from alcohol oxidation products– Comparison of daily maxima of 1-h and 8-h O3 shows a

consistent positive bias using BEIS-3 at all observed locations• However, not much bias seen based upon monitor location

type (Urban/Rural/Suburban)

31



MCNC

Future Work

l Continue IRR-type analyses

l Use actual deposition velocities for MEOH/ETOH in MAQSIP instead of PAR surrogates

l Continue evaluation using further updates to BEIS-3

l Perform sensitivities in other episodes/domains

32



MCNC

Acknowledgments

l North Carolina Department of Air Quality

Website for Results

l http://www.emc.mcnc.org/projects/NCDAQ/PGM/ results/basecase96/gdk96/index.html

Documents

Impacts of Biogenic Emissions Estimates from BEIS-3 on ...€¦ · Impacts of Biogenic Emissions Estimates from BEIS-3 MCNC The Impacts of Biogenic Emissions Estimates from BEIS-3